One of the craziest days I ever had at our library was one of our patrons brought a pony inside the library. Really. It was not a service animal. It was just a pony. She just thought the library staff would like to take a look at the pony. So she led the pony right up to the circulation desk to show the staff. A polite comment and a rather firm suggestion that ponies are better outside than inside quickly ended the matter with no hurt feelings. (Well maybe the pony didn't like this, but I didn't ask.)
Our staff, after having a few chuckles at the oddity of this, started discussing library policies. Is a pony policy necessary? One pony incident does not make for a massive pony invasion that must be stopped. Enter the problem of the "trend of one".
Pony example aside. There are other trends of one that can hurt library service. One of my regulars a few years ago finally got a colicky kid to sleep and wasn't going to shut off the engine and wake a kid just so she could run in and pick up her library holds. She asked me to run it out to the car. No big deal. I did. She thought our library was the greatest. 2 minutes of "extra" work translated into a big public relations payoff.
I brought this up to another librarian at a conference once and she was stupefied that anyone would run stuff out to the car. "Won't everyone expect that now?" Again, she was trapped by the trend of one. One time, someone asked for a little extra assistance. The world didn't stop. Library service continued without a hitch.
Assuming a trend where none exists can shut down opportunities for improving library service. Collect data and make sure decisions and rules are promoting library service.